In the most general sense, this project has been concerned with the investigation of the variables which determine the degree to which stimuli come to influence the behaviors with which they are associated. For the most part, stimulus control has been operationally defined either as stimulus generalization slope or as discriminative performance, and research has involved experimental manipulations postulated to infuleunce these measures of performance. It is expected that research will continue along many of the lines previously explored, although there will be come changes of emphasis. The role of the nature of the response requirement will be explored, as well as the transfer of an attentional set across motivation-reward systems. The assumption in the general attention postition that animals can learn generally about the significance of stimuli is consistent with another cognitive interpretation, i.e., that Ss can learn about the significance of their behavior, i.e., about the contingencies which exist between their behavior and important environmental events. We have used autoshaping to demostrate effects on acquisition of prior experience with instrumental contingencies. Further work will explore the role of amount and type of prior experience in autoshape acquisition, as well as investigating more generally the parameters which affect performance in the autoshape paradigm.